The CNS Project of ASI for Modern Celestial Navigation in Maritime

Autonomous maritime navigation presents critical challenges when operating in GNSS-denied or GNSS-degraded environments, such as coastal areas with multipath effects, industrial ports with occlusion, or strategic regions exposed to spoofing and jamming. In these conditions, traditional INS-based navigation suffers from drift accumulation, compromising long-term accuracy. In alternative, SINS/CNS systems are proving to be as accurate, autonomous, reliable, inexpensive and virtually independent of all external inputs. New automatic celestial navigation systems are now possible in the marine environment, based on the aerospace experience in star-trackers. Existing techniques relying on GNSS fusion or loosely coupled celestial systems are insufficient in guaranteeing autonomy and reliability in adverse scenarios. Furthermore, star trackers designed for space applications require substantial adaptation for marine operations, where visibility is intermittent, environmental conditions are variable, and dynamic motions affect sensor alignment. To cope with these challenges, this paper presents a novel architecture for robust and accurate navigation in GNSS-denied maritime environments, combining a tactical-grade INS with multispectral star tracker (ST) observations, embedded in a Factor Graph Optimization (FGO) framework. The project targets a FGO-based CNS system, which will ensure accurate attitude and position estimation even with partial or degraded inputs, providing a reliable backbone for maritime autonomy in GNSS-contested environments. The system will implement a resilient, drift-suppressing, and observable navigation architecture that leverages celestial sources in both day and night conditions (via SWIR), combines them tightly with inertial and Doppler sensors, and compensates for degraded observations using robust estimation techniques and AI-based error modeling.